Product-managing shaped intelligence is operationally established within the bounded AI Shaping knowledge and publication product. Under human product ownership, stewardship and decision authority, it carries recurring product-state, owner-selection, priority, sequencing, reconciliation and resumption burden while one separately governed publication capability is active at a time. The broader AI Shaping capability architecture also integrates separately governed Project-managing shaped intelligence through a singleton Launcher routing and integration surface without transferring substantive ownership.
Status: bounded operational establishment · published formal evidence · first-party, retrospective and unquantified.
The architecture remains stable while programme sequences vary. Publication co-evolution is one relationship within it; the completed sequence below is one bounded programme, not a universal order.
It carries recurring product-management burden across the AI Shaping knowledge and publication product: accepted cross-surface state, source and provenance resolution, complete-definition-owner identification, priority and dependency management, one-active-capability selection, revision sequencing, reconciliation, propagation routing, closure, state preservation and resumption.
AI shaping is the discipline; the AI Shaping knowledge and publication product is the bounded product; Product-managing shaped intelligence is the product-level coordinating capability within it. It prepares review-ready work but does not own product purpose or consequential decisions.
The complete current position is repeated accepted-cycle operation within the bounded AI Shaping knowledge and publication product. The dated table below is a selected public trace of the first four accepted end-to-end cycles; it is not the complete current cycle history. Correct owner selection, accepted-state preservation, one-active-capability routing, human acceptance, reconciliation and resumption are visible first-party operational signals.
The satisfied maturity gate and published paper do not establish quantified burden reduction, measured productivity improvement, independent validation, autonomy, generic transferability, universal Product-management capability, complete Scrum practice, deployment assurance or superiority of the direct-shaping approach.
Governed automation means recurring product-state and coordination work is carried through accepted controls and presented for review. Human product ownership retains purpose, value, audiences, priorities, scope, architecture decisions and acceptance; stewardship retains continuity and integrity; decision authority retains judgement, correction, rejection, disclosure, publication, deployment and consequential action.
This is not autonomous product management, a conventional Scrum Team, an AI employee team or authority delegated to AI.
| Cycle | Accepted sequence | Operational signal |
|---|---|---|
| 1 | Website-to-knowledge propagation | Owner selection, separate surface authority, human acceptance and receiving-surface outcome |
| 2 | Private v1.41 → website v1.18 → knowledge v1.12 harmonisation | Cross-surface priority, sequence, state preservation and reciprocal refinement |
| 3 | Website v1.19 → knowledge v1.13 assessment correction | Bounded correction, resumption, stopping discipline and gate evaluation |
| 4 | Product v1.2 → private v1.42 → website v1.20 → knowledge v1.14 product-ownership harmonisation | Owner-first propagation, cross-surface acceptance, shared-diagram adoption and end-to-end closure |
The predeclared three-cycle numerical and qualitative requirements were met before cycle 4, so the three-cycle maturity gate remains satisfied. The reconstruction and routing-burden signal remains observable, first-party and unquantified.
The complete current status is supported across repeated accepted cycles. The table remains a selected historical public trace and does not advance automatically when private product state changes.
The capability architecture is broader than the four-capability publication subsystem. Product-managing shaped intelligence coordinates the three current publication capabilities shown below. Separately governed Project-managing shaped intelligence is established through the mediated-shaping approach using AI-shaping intelligence, and Product and Project routing is integrated through one singleton Launcher while substantive ownership remains separate.
| Capability | Principal work object | Public inspectability |
|---|---|---|
| Product-managing shaped intelligence | Product purpose context, accepted cross-surface state, owner selection, priority, sequence, dependency, reconciliation and resumption across the publication product. | Its bounded role, evidence status, operating relationships and accepted outcomes may be explained publicly. Controls, prompts, source-resolution machinery, assessment and launcher administration remain private. |
| Private knowledge-publishing shaped intelligence | The private Knowledge surface, including the complete practitioner doctrine, digestible synthesis, repeated reading, preparation and change incubation. | Its name, private status and bounded purpose may be described. Content and operating machinery remain private. |
| Published paper-authoring shaped intelligence | The formal three-paper set: Category Definition, Project-managing Capability Evidence and Product-managing Capability Evidence. | Published papers, versions, diagrams, references and DOI records are inspectable; revision controls remain private. |
| Public web-developing shaped intelligence | The explanatory, navigational, visual and technical public website across authorised revisions. | The website and visible revision effects are readily inspectable; internal shaping and governance machinery remain private. |
Within the Product publication subsystem, Product-managing shaped intelligence uses publication co-evolution through reciprocal refinement across the three current separately governed publication capabilities. An improvement may originate in any surface, but the complete-definition owner resolves it before a separately authorised receiving capability propagates the accepted expression.
Publication co-evolution is not the umbrella capability, a fourth publication capability, a substitute name for Product-managing shaped intelligence, automatic propagation, simultaneous multi-agent operation or the complete product-management function.
Human product ownership establishes purpose, value and the decision boundary. The human materially finalises scope, priorities, disclosure limits and the acceptance standard.
Product-managing shaped intelligence identifies the complete-definition owner. It resolves accepted state and selects one governed publication capability for revision.
The selected publication capability revises its own work object. Every other capability remains read-only context; no surface loses its authority.
Human review accepts, corrects or rejects the result. Judgement, disclosure, publication and consequential action remain human-owned.
The accepted result returns through reconciliation. The product state records what changed, what remains deferred and whether another receiving capability is warranted.
The standing capability architecture is enduring; publication co-evolution is one operating relationship within it, and this completed programme is one bounded temporal sequence operating through that architecture. Product-managed work is initiated through planning reconciliation only when the purpose, materiality, boundaries and owner are sufficiently clear.
Discovery and Closure are programme boundary states, not maintained sets, shaped capabilities or phases of the separate ten-phase AI Shaping capability lifecycle. Discovery may produce no programme. Closure creates no automatic continuation; another programme requires separate planning reconciliation.
Maintained set: Product. Established the cross-surface contract, ownership rules, dependencies and stage gates.
Accepted outcome: Product v1.17.
Acceptance-reconciliation stage gate: human review and acceptance closed the phase before another capability was considered.
Maintained set: Knowledge. Provided the exhaustive private operating-model explanation and state-reacquisition treatment.
Accepted outcome: Knowledge v1.35.
Acceptance-reconciliation stage gate: accepted Knowledge work was reconciled before Paper work was activated.
Maintained set: Papers. Formalised the programme's bounded evidentiary significance while preserving the public/protected boundary.
Accepted outcome: internal Papers v1.91, unpublished.
Acceptance-reconciliation stage gate: internal Paper acceptance did not itself publish or propagate the work.
Maintained set: Private. Aligned doctrine, strategic decisions, concordance and implementation orientation within the private boundary.
Accepted outcome: Private v1.52.
Acceptance-reconciliation stage gate: Private acceptance preceded any separately justified public expression.
Maintained set: Website. Presented the bounded programme publicly without exposing protected operating machinery.
Accepted outcome: Website v1.37.
Programme-closure gate: human acceptance closed the programme after no further material and uniquely owned corrective phase remained.
Every activated capability used a governed release cycle. Every candidate received human review and acceptance, acceptance reconciliation closed each phase, and a separate stage gate activated the next phase only when the work remained material, source-complete, uniquely owned and unambiguous. Exactly one shaped capability was active at a time.
Complete-definition ownership, accepted/current and rollback identities, correction and recovery, resumption from accepted state, explicit non-action and prevention of automatic propagation remained visible. Optional, ambiguous and non-material propagation was stopped or deferred. Final programme Closure occurred on 30 July 2026.
Role allocation: Product owns the cross-surface contract and programme coordination; Knowledge provides the exhaustive private explanation; the Website presents one bounded public-safe instance; the Product-managing Capability Evidence paper formally interprets its evidentiary significance; and Knowledge retains complete practitioner-doctrine and private explanatory roles within its private boundary; terminal retired Private remains historical/rollback/reconstruction evidence only.
Evidence boundary: this is first-party, retrospective, source-bounded and unquantified evidence without a controlled baseline, causal or controlled comparative effect, independent validation or replication. It does not establish generic transferability, autonomy, complete Scrum practice, universal product-management capability, deployment assurance, implementation-transfer authority or Product-managing cycle 18. The explanatory frame adds no new empirical evidence.
The example is one bounded temporal instance of the standing architecture, not a universal five-phase lifecycle, permanent release train, operating manual or proof that every programme requires the same capabilities or order. The later outcome and presentation-assurance programme is separate and does not reopen or renumber this completed programme.
The capabilities do not act concurrently or maintain permanent shared awareness. Sequential activation preserves versioning, source authority, disclosure decisions and clear stopping conditions.
An accepted result returns to the product state. A dependent receiving capability is selected only when one next route is unambiguous; otherwise the cycle stops.
Project-managing shaped intelligence and Product-managing shaped intelligence remain peer domain-facing capability types with different work domains, reuse boundaries, shaping histories and evidence maturity. Project-managing is the formally published mediated-shaping capability-evidence instance and is established through AI-shaping intelligence; Product-managing is separately published as a bounded direct-shaping product-level evidence instance.
The singleton Launcher integrates their routing surfaces without becoming a substantive capability or transferring ownership. It is alias-only, non-executing and non-substantive: Product and Project retain their own substantive meanings and controls, while human ownership, stewardship and decision authority remains above acceptance and consequential action. Their evidence objects, shaping approaches, limitations and conclusions remain distinct.
| Inspectable result | What it supports | What it does not establish |
|---|---|---|
| Accepted product controls and state architecture | Governed capability basis | Generic transferability |
| Public evidence snapshot: four accepted cycles | Bounded operational establishment | Quantified productivity improvement |
| Correct owner and capability routing | Product-management coordination burden carried | Independent validation |
| Human review and acceptance | Governed automation under human authority | Autonomous product management |
| Formal-paper separation | Clear evidence and publication boundaries | Generic or independently validated Product-management capability |
Show the mechanism and observable result; protect the machinery. Prompt bodies, source-resolution procedures, shaped controls, detailed reconciliation logic, validation and packaging machinery, rubric internals, launcher administration and sufficient implementation-transfer detail remain protected.
The separate AI Shaping product ownership page translates this bounded operating evidence into product-ownership, product-management, product-operations and AI-enabled delivery language. It does not create another capability-evidence role or expand either published evidence paper beyond its own evidence object.
This page supports bounded public-stage evaluation of Product-managing shaped intelligence and its publication-co-evolution mechanism. It does not disclose protected implementation material, provide deployment assurance, claim complete Scrum practice or authorise implementation or commercial use. Project-managing shaped intelligence remains the formally published mediated-shaping capability-evidence instance; the Product-managing paper is a separate bounded product-level result.
Page 17 of 31